Asphalt Property Prediction Using Virtual Cut Point
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for predicting the properties of asphalt blends from multiple crude sources are unreliable, leading to conservative feed blends and reduced distillation throughput, as they fail to accurately characterize the quality of asphalt fractions, necessitating extensive storage and slowing refinery processes.
Innovation Solution
A method involving the measurement of kinematic viscosity at 100° C. to 150° C. to determine a virtual cut point for a virtual asphalt blend, allowing for the calculation of other properties and real-time adjustment of the cut point temperature to achieve desired specifications without laboratory characterization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional experimental characterization methods are used to determine asphalt specifications, then measurement precision is improved, but productivity deteriorates due to extensive storage requirements and process delays
Solution Approach 1:
The patent replaces complex mechanical/chemical experimental characterization systems with a computational prediction system. A neural network model predicts asphalt specifications (penetration, softening point, viscosity) from crude oil blend composition and distillation conditions, eliminating the need for extensive physical testing and storage while maintaining measurement precision
Solution Approach 2:
The patent performs preliminary computational characterization of asphalt properties before the actual distillation and blending processes are completed. By using the neural network to predict specifications in advance based on feed composition and cut points, the system determines asphalt grade and suitability prior to manufacturing, avoiding post-production delays
2Reliability
If conservative feed blends with higher percentage of heavy asphaltic feeds are used, then reliability of meeting specifications is improved, but productivity deteriorates due to limited distillation throughput
Solution Approach 1:
The patent changes the approach from adjusting physical feed composition conservatively to optimizing computational parameters. The neural network model allows precise prediction of asphalt properties for any given blend composition and distillation cut point, enabling the selection of optimal (not conservative) feed blends that maximize throughput while guaranteeing specification compliance through accurate prediction
Solution Approach 2:
The patent implements a feedback mechanism where the neural network prediction system continuously evaluates proposed feed blends and distillation parameters against specification requirements. This feedback loop allows real-time optimization of blend composition and cut points to achieve maximum throughput while ensuring specifications are met, eliminating the need for conservative over-blending
3Measurement precision
If extensive storage tanks are used to hold asphalt during characterization, then measurement precision is improved, but loss of time increases due to storage requirements and process delays
Solution Approach 1:
The patent replaces the mechanical storage and physical characterization system with a computational prediction system. The neural network model instantly predicts asphalt specifications from process parameters, eliminating the time-consuming storage and experimental testing phases while maintaining accurate specification determination
Solution Approach 2:
The patent creates a virtual copy of the asphalt product through computational modeling. Instead of physically storing and testing actual asphalt, the system generates a digital representation (neural network prediction) that accurately replicates the expected specifications, allowing immediate characterization without physical handling or storage time
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables real-time prediction and optimization of asphalt properties, allowing for immediate product release, improved manufacturing efficiency, and reduced storage needs by accurately predicting asphalt quality based on viscosity measurements, thus enhancing refinery operations.
Implementation Method 1
The crude petroleum is separated into its various fractions through a distillation process
Implementation Method 2
measuring a kinematic viscosity of an asphalt fraction at a temperature of 100° C. to 150° C.
Data Source
AI summary
Methods are provided for predicting the properties of an asphalt fraction that contains two or more asphalt components based on measurements of the viscosity for the asphalt fraction. Based on the measured viscosity, a virtual cut point is determined for a virtual asphalt blend that has the same viscosity (to within a tolerance value) as the measured viscosity for the asphalt fraction. The virtual cut point is then used to determine a variety of predicted property values for the asphalt fraction. Optionally, the predicted property values can be used to adjust the actual cut point for the distillation or separation process used for forming the asphalt fraction.